Saturday, 15 August 2015

The (slight) positive relationship between teaching and research (based on NSS and REF data for law)

Ok, this is not my original idea but inspired by A. Afonso’s calculation for politics departments. Of course, the relationship between teaching and research — and the question of possible synergies or trade-offs — is also a general topic of research in higher education. And, with respect to law, a recent article by Ginsburg and Miles found that under some specifications research and teaching are positively correlated (based on data from the University of Chicago law school).
   My approach was as follows: I took the 66 law units for which data for both research (ie the REF 2014 data) and student satisfaction (ie the NSS 2015) are available. I started with the NSS data on ‘overall satisfaction’ — but the relation between those and the REF data turned out to be flat and insignificant. Thus, the following is based on the NSS teaching scores only (ie disregarding the other categories of the NSS). For the REF I started with the GPA (ie the quality score):

This shows only a slight positive correlation (0.076) which is not statistically significant.
    But there are also a number of other ways of presenting the REF data, eg, considering the number of *s, or just 3 and/or 4*s; and then scaled by the number of staff submitted or the eligible staff, or else just the absolute numbers. 
   The strongest correlation could be identified for the total number of 4*s - ie the 'research power' of law schools based on 4* assessments:

Here the correlation is 0.2534 and significant at the 5 % level. Note that this result, based on a research power indicator, combines two effects as the total number of 4*s is both dependent on the size of the law school and their respective research quality. In the chart it can be seen that the correlation is mainly due to (i) the law schools with the lowest NSS scores also having few 4* assessments and (ii) the law schools with most 4* assessments usually having above average NSS scores.
    Caveat: there are of course good reasons to be sceptical about the REF and the NSS data; but still, as both are based on subjective assessments, there may be some justification in comparing the corresponding data points.

Friday, 14 August 2015

Publications update (on themes of company and comparative law)

Publishers don’t seem to do summer breaks as in the last few weeks a couple of my articles and book chapters have been published ... So here they are: